One of the other testers and I were talking this morning about unit tests.  Here’s an example of 100% unit tests passing, 100% code coverage and still results in total failure:

http://articles.cnn.com/1999-09-30/tech/9909_30_mars.metric.02_1_climate-orbiter-spacecraft-team-metric-system?_s=PM:TECH

Essentially, the US tests used English measurements but the European team used metric. In turn, this caused a multi-million dollar spacecraft to fail.

Page 6 here of the post mortem explains the details: ftp://ftp.hq.nasa.gov/pub/pao/reports/1999/MCO_report.pdf (This is a really good post mortem analysis of what went wrong, by the way. Much of it is specific technical details that I do not understand, but it is well worth the read.)

Good reading here and lots to think about. So that's what I am doing now - thinking about this.

Questions, comments, concerns and criticisms always welcome,

John